As long as a system is democratic, based upon free and fair elections with universal suffrage, political institutional arrangements do not matter most for public policy. Where institutional arrangements restrict the size of the electorate electing policymakers from universal suffrage of all eligible adults, whom those standing for election will answer to (such as with more ad hoc internal party mechanisms to choose the leadership), they matter more for public policy than they do otherwise. However, apart from this, public preferences are broadly translated into public policy. Electoral systems—whether proportional, majoritarian, or hybrid—tend to reflect the broad direction of public opinion, even if the winning party under majoritarian systems is over-represented compared with its vote share. The largest party in a government is almost invariably the one that won the most support at election-time. Thus, elected governments broadly reflect public opinion, and other actors such as bureaucracies are subordinate to them.
Public preferences strongly influence public policy under any democratic political institutional arrangement. There is a scholarly consensus that the democratic policy-making process begins with public preferences, expressed through elections or direct democracy. Powell (2004) conceives of this process as a four-stage “chain of responsiveness”: the first stage is “Citizens’ Preferences”, moving on to “Citizens’ Voting Behavior”, “Selecting Policy Makers” and, finally, “Public Policies and Outcomes”. These four stages are connected by three consecutive linkages: “Structuring Choices”, “Institutional Aggregation” and “Policy Making”. Powell argues that any breaks in these linkages are detrimental to public preferences’ being translated into policy outcomes. He emphasises “institutionalized arrangements, above all elections” rather than merely “the good will of policy makers” in connecting voters to policymakers in democracies. It is true that politicians’ need to win votes to hold power is the most powerful institutional incentive encouraging elected policymakers to follow public preferences. Powell’s chain, in fact, loops back on itself: elected politicians must answer to voters for their actions at regular elections.
Moreover, the power of public opinion remains whatever the democratic institutional arrangements. Countries such as Switzerland frequently use direct democracy to settle policy, such as with the 2005 referendum to join the Schengen treaty (adopted) or the 2011 referendum on increased gun controls (rejected). Electoral systems also vary between majoritarian, proportional, or hybrid systems. Golder and Stramski (2010) find legislatures are more representative of public preferences under proportional systems (such as Germany’s) than under majoritarian systems (such as Britain’s). However, governments under proportional systems are not substantively more congruent with voters’ ideological preferences. Britain’s 2016 EU referendum illustrates public preferences becoming public policy. Voters voted to Leave by referendum, which became government policy, confirmed at two subsequent general elections. Even though anti-Brexit parties such as the Liberal Democrats and the SNP were emboldened by the support of some electoral groups, they did not represent the will of the majority; hence, they could not enact their policy through elections. Even with representative democracy, the policy views of the newly-elected MPs in 2017 and 2019 were shaped by public preferences.
Whilst public preferences fundamentally shape public policy in democracies, elected MPs are not formally obliged to follow the wishes of their constituents or to keep the manifesto pledges upon which they were elected under representative democracy. Thus, their personality and beliefs can have a great influence upon public policy once in power. Principal-agent theory (Strøm, 2000) stipulates that the agent (representative) whom the principal (voters) delegates to act on their behalf is not obliged to follow their wishes. Thus, when Liz Truss became Prime Minister, she proposed a radically different agenda from the manifesto upon which the Conservatives had won their electoral mandate in 2019. Her policies included abolishing the cap on bankers’ bonuses and the 45p rate of income tax on the highest earners. Truss had been a high-ranking minister in Boris Johnson’s Cabinet in 2019 as International Trade Secretary. Nonetheless, she was not formally bound by the Conservatives’ manifesto pledges.
According to principal-agent theory, MPs (agents) can choose to expound their own beliefs even when they contradict the views of their constituents or the manifesto they stood on, as power has been delegated to them by voters (the principal). However, the principal also closely monitors the agents’ actions, and can eject them at election-time if they are dissatisfied with them. Within the same institutional framework, a different leadership style can produce very different results. This highlights the role that personality and beliefs can play, even if politicians are constrained by ultimately having to answer to voters at election-time. For example, as Prime Minister and Chancellor, Rishi Sunak and Jeremy Hunt respectively pursued a far tighter fiscal policy than Truss and her Chancellor Kwasi Kwarteng had, thus reassuring the financial markets and lowering Britain’s borrowing costs. The former pair had very different beliefs and more cautious personalities than the latter pair. Sunak and Hunt believed in reassuring markets, balancing budgets and controlling inflation before cutting taxes. This underlines how party leadership election mechanisms that restrict the size of the electorate can shape public policy, bringing to power a new leader who does not necessarily have any loyalty to the party’s promises from the most recent election. In contrast, a leader elected on a set of promises will generally seek to demonstrate to their voters that they are implementing them, to bolster their chances of re-election next time.
Party leadership election mechanisms matter more for public policy than democratic national electoral systems. Instead of including the entire national electorate, these mechanisms restrict suffrage to a smaller group, such as a party’s MPs or members. Thus, these mechanisms break the linkages in Powell’s chain from “Citizens’ Preferences” through to “Public Policies and Outcomes”, as citizens’ voting behaviour does not select policymakers. Consequently, the policy preferences of that smaller group matter most to candidates standing for office, rather than those of all citizens eligible to vote in national elections.
They also underline the agenda-setter’s power to shape political outcomes by constraining choices (Clark, Golder and Golder, 2017). For example, the 1922 Committee, representing Conservative backbenchers, sets the rules of Conservative leadership elections: whether Conservative members or only MPs can vote, the nominations required to stand, and the rounds of voting. This affects which candidate wins and shapes public policy, as in the cases of Truss and Sunak. The latter won among MPs but lost among party members in 2022. MPs can vote strategically, lending an ‘insincere vote’ to a candidate who would not be their first choice so that the candidate they favour the least is eliminated in an early round of voting.
In the key executive-legislative relationship under Britain’s parliamentary system, King (1976) sees government backbenchers as needing government policies they support and can sell at election-time. Meanwhile, the government needs its backbenchers’ parliamentary votes and their public support. Since King’s article, the breakdown of party discipline, particularly within the Conservative Party, means government backbenchers are more likely to rebel on votes—such as over the Syrian military intervention in 2013—or oust the party leader. The last three Conservative Prime Ministers—Theresa May, Johnson, and Truss—were all removed from office by Conservative MPs. This contrasts with a presidential system (such as in the United States or France), with a separation of powers between the executive and legislature. Under such a system, the former does not depend upon the latter for its political authority. There are means of removing the President, such as impeachment. However, these are much harder to enact, are used much more rarely, and require cross-party. Thus, an unpopular incumbent is much more likely to be removed in Britain than in France or the United States.
Whatever the mechanism by which a democratic political leader has been selected, the bureaucracy is subordinate to the executive, whence its authority is delegated (Müller, 2011). Hence, in countries with either permanent or political civil services, it will always matter less than the executive. Where the civil service is politically appointed, appointees are named to carry out the wishes of their political masters. Where the civil service is permanent and independent, its role is to implement the policies of the incumbent government within the law. Huber’s (2000) concepts of ex ante (“police patrol”) and ex post (“fire alarm”) strategies for executive oversight of the bureaucracy underline this subordination. Whether ministers exert extensive prior control over civil servants’ actions or monitor them with a mixture of “carrots” and “sticks”, ministers’ ability to do so shows that civil servants’ authority is delegated. If bureaucracies were more powerful than governments, changes of the governing party would not produce significant policy changes. In practice, successive governments implement radically different agendas, as with the change from James Callaghan’s Labour government to Margaret Thatcher’s Conservative government in 1979. Since newly-elected governments broadly reflect public preferences at the time of their election, public preferences ultimately have more sway over public policy than bureaucracies.
The literature finds that bureaucratic agencies are less likely to act on policy signals from ministers than are cabinet-level departments. They are further removed from ministerial control and, consequently, have less direct contact with ministers (Wonka and Rittberger, 2010; Egeberg and Trondal, 2009). Governments often judge that independent regulatory institutions, notably central banks, are able to make better policy over time, free from party-political considerations. The consistency and credibility of such institutions with relevant actors, such as with central banks and financial markets, create an environment beneficial for desired outcomes. However, the executive has the power to revoke their independence. Furthermore, the executive can continue to exert indirect control, as with the Chancellor of the Exchequer’s power to appoint four members of the Bank of England’s Monetary Policy Committee. This underlines how, within the framework of democracy, the power of the democratically-elected executive and legislature means that public preferences are generally translated into public policy.
In conclusion, political institutional arrangements do not matter most for public policy: other variables can change public policy under the same institutional structures. Overall, public opinion is reflected in elections under majoritarian and proportional systems, with governments reflecting public preferences very similarly. Politicians’ personalities and beliefs can also shape public policy independently of institutions, albeit within the constraint of answering to public opinion at elections. Whether civil services are politically appointed or permanent and independent, bureaucracies are subordinate to elected governments, implementing their policies. Even delegated bureaucratic authority over policy, such as with central banks, is revokable. Institutional arrangements can be influential when they restrict the size of the electorate and control the electoral process (such as nominations required and rounds of voting), as with party leadership elections. This helps shape public policy by determining which candidates can stand and are best-placed to win. Clearly, other political institutional arrangements still matter for public policy. Bureaucracies’ role in implementing policy matters for the policy’s efficiency, even if the government or legislature’s role in dictating policy matters more. Public opinion is not perfectly reflected in governments produced by elections (few voters ever agree with every policy of their party, for example) and electoral systems influence legislatures’ representativeness. However, governments ultimately reflect most voters’ preferences among the electoral offerings.
Clark, W.R., Golder, M. and Golder, S.N. (2017) ‘Problems with group decision making’, in W.R. Clark, M. Golder and S.N. Golder, Principles of comparative politics. 3rd edn. Washington, DC: SAGE/CQ Press.
Egeberg, M. and Trondal, J. (2009) ‘Political leadership and bureaucratic autonomy: effects of agencification’, Governance: An International Journal of Policy, Administration, and Institutions, 22(4), pp. 673-688.
Golder, M. and Stramski, J. (2010) ‘Ideological congruence and electoral institutions’, American Journal of Political Science, 54(1), pp. 90-106.
Huber, J.D. (2000) ‘Delegation to civil servants in parliamentary democracies’, European Journal of Political Research, 37(3), pp. 397-413.
King, A. (1976) ‘Modes of executive-legislative relations: Great Britain, France and West Germany’, Legislative Studies Quarterly, 1(1), pp. 11-36.
Müller, W.C. (2000) ‘Political parties in parliamentary democracies: making delegation and accountability work’, European Journal of Political Research, 37(3), pp. 309-333.
Powell, G.B. (2004) ‘The Chain of responsiveness’, Journal of Democracy, 15(4), pp. 91-104.
Rittberger, B. and Wonka, A. (2010) ‘Credibility, complexity and uncertainty: explaining the institutional independence of 29 EU agencies’, West European Politics, 33(4), pp. 730-752.
Strøm, K. (2000) ‘Delegation and accountability in parliamentary democracies’, European Journal of Political Research, 37(3), pp. 261-289.
©George Alexander Charles. This article is licensed under a Creative Commons Attribution 4.0 International Licence (CC BY).